On the fundamental understanding of divide-and-conquer methods